Antitrust Chronicle SEPTEMBER VOLUME 3 SUMMER 2018 - COMPETITION POLICY INTERNATIONAL

Page created by Kristen Perkins
 
CONTINUE READING
Antitrust Chronicle SEPTEMBER VOLUME 3 SUMMER 2018 - COMPETITION POLICY INTERNATIONAL
Antitrust
  Chronicle
SEPTEMBER · VOLUME 3 · SUMMER 2018

     PLATFORM COMPETITION
                                     COMPETITION POLICY
                                     INTERNATIONAL
Antitrust Chronicle SEPTEMBER VOLUME 3 SUMMER 2018 - COMPETITION POLICY INTERNATIONAL
TABLE OF CONTENTS
03                                             26
     Letter from the Editor                         Public Interest Journalism, the Internet,
                                                    and Competition for Advertising
                                                    By Henry Ergas, Jonathan Pincus & Sabine
                                                    Schnittger

04                                             32
     Summaries                                      Platform Power and Privacy Protection: A
                                                    Case for Policy Innovation
                                                    By Caron Beaton-Wells

06                                             40
     What’s Next?                                   The Tragedy of the Successful Firm
     Announcements                                  By Konstantinos Stylianou

07                                             46
     CPI Talks…                                     Online Platforms and Antitrust: Evolution
     …with Thomas Kramler                           or Revolution?
                                                    By Renato Nazzini

10                                             52
     Strengthening Buyer Power as a Solution        Two-Sided vs. Complementary Products
     to Platform Market Power                       By Lapo Filistrucchi
     By Hugh Mullan & Natalie Timan

18
     An Introduction to the Competition Law
     and Economics of “Free”
     By Benjamin Edelman & Damien Geradin
Antitrust Chronicle SEPTEMBER VOLUME 3 SUMMER 2018 - COMPETITION POLICY INTERNATIONAL
Editorial Team
CEO & Founder
                                           LETTER FROM THE EDITOR
David S. Evans                             Dear Readers,
Editor in Chief
Elisa V. Mariscal                          It has been fifteen years since Rochet & Tirole published their influential article Platform
                                           Competition in Two-Sided Markets. The September 2018 CPI Antitrust Chronicle features
Managing Director
Raúl Escalante                             articles on today’s debates related to Platform Competition.

Managing Editor,                           We are delighted to open the September Chronicle with an interview with Thomas Kram-
CPI Projects
Elisa Ramundo                              ler, Head of the Digital Single Market Task Force in the European Commission's Direc-
                                           torate General for Competition. Before the year closes out, we expect to also have an
Managing Editor,                           interview (same questions) with Bruce Hoffman, Director of the Bureau of Competition
The Antitrust Chronicle
Sam Sadden                                 at the U.S. Federal Trade Commission.

Senior Editor                              As the platform economy continues to evolve and grow, how should regulators, stake-
Nancy Hoch
                                           holders, and the rest of us evolve and adapt? For instance, as one of this month’s authors
Latin America Editor                       points out, “[a]s economies and societies continue to be transformed by the data revo-
Jan Roth                                   lution, privacy protections will continue to be paramount.” Will other factors remain or
Junior Editor                              become paramount? Will the trustbusters of tomorrow and their tools resemble those of
Jeff Boyd                                  yesteryear, or will they be a completely new breed?

Editorial Advisory Board                   We are also pleased to announce CPI’s inaugural annual conference on “Challenges
                                           to Antitrust in a changing economy” that will take place on November 9, 2018 at
Rosa Abrantes-Metz                         Harvard Law School. This event, co-organized with CCIA, aims to generate an open and
Stern School of Business
                                           cutting-edge debate on competition law and economics in the tech industry. Confirmed
Kent Bernard                               speakers include: Bruce Kobayashi, William Kovacic, Einer Elhauge, Hal Varian, Diana
Fordham School of Law                      Moss, Nancy Rose, Julie Brill, Greg Ip, Jonathan Baker, Rob Atkinson, and Michael Man-
Rachel Brandenburger
                                           del, among others. Read the full program and register free online here.
Oxford University
                                           Lastly, please take the opportunity to visit the CPI website and listen to our selection of
Dennis W. Carlton                          Chronicle articles in audio form. A growing number of subscribers are getting their CPI fix
Booth School of Business
                                           on the go, rather than in the office; audio is an excellent option for this!
Adrian Emch
Hogan Lovells                              As always, thank you to our great panel of authors.
Kyriakos Fountoukakos
Herbert Smith                              Sincerely,
Jay Himes
Labaton Sucharow                           CPI Team1

James Killick
White & Case

Stephen Kinsella
Sidley Austin

Ioannis Lianos
University College London

Robert O’Donoghue
Brick Court Chambers

Aaron Panner                               1 CPI thanks CCIA for their sponsorship of this issue of the Antitrust Chronicle. Sponsoring an
Kellogg, Hansen, Todd, Figel & Frederick   issue of the Chronicle entails the suggestion of a specific topic or theme for discussion in a given
                                           publication. CPI determines whether the suggestion merits a dedicated conversation, as is the
Vanessa Yanhua Zhang                       case with the current issue of the Chronicle. As always, CPI takes steps to ensure that the view-
Renmin University                          points relevant to a balanced debate are invited to participate and that the quality of our content
                                           maintains our high standards.

                                                            3                                            CPI Antitrust Chronicle September 2018
Antitrust Chronicle SEPTEMBER VOLUME 3 SUMMER 2018 - COMPETITION POLICY INTERNATIONAL
SUMMARIES
  07                                                                                      10

CPI Talks…                                                                              Strengthening Buyer Power as a Solution to Platform
                                                                                        Market Power
…with Thomas Kramler
                                                                                        By Hugh Mullan & Natalie Timan
In the month’s edition of CPI Talks… we have the pleasure of speaking with
Thomas Kramler. Mr. Kramler is Head of the Digital Single Market Task Force             There are numerous examples of digital platforms with market power. There
in the European Commission's Directorate General for Competition.                       have been calls to regulate these platforms. This article explores an alterna-
                                                                                        tive way of tempering any significant market power by changing the relative
        Mr. Kramler addresses questions related to the digital economy, how             bargaining strength of the platform vis-à-vis its customers. The bargaining
well do existing competition rules work in digital markets, and the pressures           strength of any customer will depend on the number and strength of the
on competition rules to increasingly absorb public policy objectives such as            credible outside options available to that customer. Platforms, which are
privacy or data protection.                                                             multi-sided in their nature, display network effects. These network effects, in
                                                                                        turn, create coordination problems that make it harder for the customers of
                                                                                        platforms to switch to outside options. In this way, the potential exercise of
                                                                                        buyer power is more limited. Collective bargaining might be one way to solve
                                                                                        this coordination problem and in doing so, provide customers with greater
                                                                                        bargaining strength. Such collective bargaining will only benefit consumers if
                                                                                        the coordination to create or strengthen outside options does not give rise to
                                                                                        competition concerns through coordination.

  18                                                                                      26

An Introduction to the Competition Law and                                              Public Interest Journalism, the Internet, and
Economics of “Free”                                                                     Competition for Advertising

By Benjamin Edelman & Damien Geradin                                                    By Henry Ergas, Jonathan Pincus & Sabine Schnittger

Many of the largest and most successful businesses today rely on providing              The advent of the Internet and of new Internet-based platforms and social
service at no charge to at least a portion of their users. Free services often          networks has profoundly changed the competitive landscape for traditional
delight users, yet also create a series of challenges for competition policy, in-       publishers of “public interest journalism” or, more generally, quality journal-
cluding impeding entry, inviting overproduction on quality, and increasing the          ism. Consumers now have access to virtually unlimited online content, of
risk of deception and overpayment. This short paper presents these problems,            which quality journalism is one of many options. But despite clear gains for
examines the strategies that entrants can attempt when competing with free              consumers, there has been widespread concern about the long run impacts
service, and considers possible regulatory responses.                                   of the changes now underway on the future of the medium, and of journal-
                                                                                        ism more broadly. In this article, we investigate how recent trends towards
                                                                                        digitalization have affected the provision of public interest journalism in Aus-
                                                                                        tralia, how they might play out in the future and the scope for public policy
                                                                                        responses.

                                                                                    4                                             CPI Antitrust Chronicle September 2018
Antitrust Chronicle SEPTEMBER VOLUME 3 SUMMER 2018 - COMPETITION POLICY INTERNATIONAL
SUMMARIES
  32                                                                                        40

Platform Power and Privacy Protection: A Case for                                         The Tragedy of the Successful Firm
Policy Innovation
                                                                                          By Konstantinos Stylianou
By Caron Beaton-Wells
                                                                                          Twenty years after the emergence of platform literature antitrust courts and
Are the goals of competition and privacy in alignment or in conflict? The                 authorities still face formidable challenges in deciphering platform business
antitrust-privacy interface is the subject of much academic and policy debate,            models. The latest two Google cases (Google Search, Google Android) high-
particularly in the context of digital platforms. This article maps the contours          light the schism between what firms consider an innovative business plan
and underlying values of two models for managing the interface, one of policy             and what the law is ready to accept. Noting the schism, this article focuses
consistency (as manifested in Europe) and one of policy separation (as mani-              on a number of points that European competition law needs to address to
fested in the United States). It identifies a third model, one of policy innovation       better reconcile business practice with antitrust legal analysis with the view
(as manifested in Australia), and explains why this model has the potential to            to enable successful firms to operate without fear of punitive enforcement.
overcome limitations inherent in alternatives.

  46                                                                                        52

Online Platforms and Antitrust: Evolution or                                              Two-Sided vs. Complementary Products
Revolution?
                                                                                          By Lapo Filistrucchi
By Renato Nazzini
                                                                                          In his dissenting opinion in Ohio et al. vs American Express, Justice Breyer
This article asks the question as to whether existing antitrust tools are suit-           seems to call for a clarification, from Filistrucchi et al. (2014) on the dis-
able for dealing with perceived antitrust concerns raised by online platforms             tinction between products sold by two-sided platforms and complementary
by focusing on three examples: (1) market definition and market shares; (2)               products. The question however has been lingering also in economic circles.
barriers to entry; and (3) the role of innovation as a safe harbor or as a de-            In fact, it has often been the case in the past years that colleagues working in
fense. It concludes that these concepts should and will continue to play a sig-           different fields have (often wrongly, but not always so) commented to me that
nificant role in the antitrust analysis of online platforms but need to evolve and        two-sided markets are like markets for complement products. Without tak-
adapt to the new requirements of the digital economy. Costly mistakes, both               ing any stance on whether the alleged behavior by American Express should
in terms of failure to intervene when required and erroneous interventions,               or should not have been deemed illegal, I discuss why and to what extent
may occur if this does not happen at the speed that the pace of economic                  two-sided platforms are different from platforms selling complement prod-
development requires.                                                                     ucts. I also explain why the distinction is relevant in assessing firms’ behavior
                                                                                          for the purposes of competition policy.

                                                                                      5                                             CPI Antitrust Chronicle September 2018
Antitrust Chronicle SEPTEMBER VOLUME 3 SUMMER 2018 - COMPETITION POLICY INTERNATIONAL
WHAT’S NEXT?
Our October 2018 Chronicle will feature articles from speakers at the CRESSE Conference which took place in Crete this summer.

ANNOUNCEMENTS
CPI wants to hear from our subscribers. In the remaining months of 2018, we will be reaching out to members of our community for your feed-
back and ideas. Let us know what you want (or don’t want) to see, at: antitrustchronicle@competitionpolicyinternational.com.

CPI ANTITRUST CHRONICLE NOVEMBER 2018 & DECEMBER 2018

The November 2018 Chronicle will focus on issues related to Due Process & Antitrust.

The December 2018 Chronicle will have articles that address issues related to Multi-Sided Markets & Consumer Harm and notably the recent
Google Android decision in the EU.

Contributions to the Antitrust Chronicle are about 2,500 – 4,000 words long. They should be lightly cited and not be written as long law-review
articles with many in-depth footnotes. As with all CPI publications, articles for the CPI Antitrust Chronicle should be written clearly and with the
reader always in mind.

Interested authors should send their contributions to Sam Sadden (ssadden@competitionpolicyinternational.com) with the subject line “Antitrust
Chronicle,” a short bio and picture(s) of the author(s).

The CPI Editorial Team will evaluate all submissions and will publish the best papers. Authors can submit papers in any topic related to competi-
tion and regulation, however, priority will be given to articles addressing the abovementioned topics. Co-authors are always welcome.

                                                                         6                                        CPI Antitrust Chronicle September 2018
Antitrust Chronicle SEPTEMBER VOLUME 3 SUMMER 2018 - COMPETITION POLICY INTERNATIONAL
CPI TALKS…

…With Thomas Kramler

In this month’s edition of CPI Talks… we have the pleasure of speaking with Thomas Kramler. Mr. Kramler is Head of the Digital Single Market
Task Force in the European Commission’s Directorate General for Competition.

Thank you, Mr. Kramler, for sharing your time for this interview with CPI.

1. Is the current antitrust framework well suited to deal with harm to innovation, especially in the context of the digital economy?
   Or put another way, how well do existing competition rules work in digital markets? Is there a need to adapt the rules?

The debate over whether competition law is capable of dealing with new market developments in innovative markets is as old as competition
law enforcement. In 1958, when the competition provisions of the Treaty of Rome entered into force, the skateboard, and more importantly,
microchips, which are today used in virtually every piece of electronic equipment, had just been invented. Evidently, the drafters of the Treaty of
Rome could not have foreseen the technical developments that followed, such as the smartphone. However they wisely formulated Articles 85
and 86 of the Treaty (now 101 and 102 TFEU) in a manner that allows them to take into account technical developments and even turn them into
a yardstick for the assessment of restrictions to competition under EU competition law.1

        Every market is specific. The basic tools of European competition law (Articles 101 and 102 TFEU) have been crafted in a “technology
neutral” way and can be applied to “new” or “old” markets alike. The European Court of Justice confirmed in Telia Sonera2 that the application
of EU competition rules cannot depend on whether the market concerned has already reached a certain level of maturity and that particularly in
rapidly growing markets, quick intervention may be warranted to prevent harm to competition.

        The history of EU competition law enforcement since 1958 shows that the basic EU competition rules are flexible enough to address
technological developments. This is well illustrated by the examples of the IBM abuse of dominance case on mainframe interface information
in the 1980s, which was settled in 1984, and the Microsoft interoperability abuse of dominance case, which was decided two decades later in
2004. Both cases addressed technological challenges of their time (mainframe and work group server interoperability) within the legal framework
set out by the Treaty of Rome in 1958.

        In the Microsoft case, the question of whether its refusal to supply interoperability information limited technical development, i.e., innova-
tion by competitors in work group servers, was actually at the heart of the case and addressed extensively in the Commission’s 2004 decision
and the 2007 Court judgment. The case, which led to the emergence of an innovative open source competitor to Microsoft in the work group
server operating system market, shows that EU competition law can effectively be applied to cases which concern innovation in technology
markets.

         EU competition rules as well as their application by the EU courts have therefore proved flexible enough to address technological develop-
ments. This does not, however, mean that antitrust enforcers should be complacent about new market developments. On the contrary: in order to
remain relevant, the established enforcement principles need to be applied in the context of market realities and, where necessary, adapted in the
light of these realities. Market features of digital markets, such as the growing importance of data and network effects and the provision of “free”
services to consumers in two sided digital platform markets, can and must be factored into the assessment of conduct under EU competition law.
The Commission’s Google Shopping and Google Android decisions of 2017 and 2018 demonstrate this.
1 See the references to the “promotion of technical progress” in Article 101(3) TFEU and to “limiting the technical development to the prejudice of consumers”
in Article 102 (b) TFEU.
2 See C‑52/09, TeliaSonera Sverige, para. 108.

                                                                             7                                           CPI Antitrust Chronicle September 2018
Antitrust Chronicle SEPTEMBER VOLUME 3 SUMMER 2018 - COMPETITION POLICY INTERNATIONAL
2. Do you think that the transatlantic antitrust differences, if any, will grow even more in light of the digital economy?

One should not overstate transatlantic differences when it comes to competition law enforcement. The overall goal, which is to protect compe-
tition in the interest of consumers, is shared across the Atlantic and there is a lot of very close cooperation between U.S. and EU enforcers and
convergence in their approaches and outcomes for the vast majority of cases.

        However, when it comes to monopolization/abuse of dominance cases, one has to acknowledge that the law on both sides of the Atlantic
is somewhat different. Enforcement priorities might also differ, including in digital markets, and first and foremost, market conditions may be
different. This can lead to different assessments of the need to intervene in certain markets.

        More generally, I would say that the EU approach towards the actions of companies with considerable market power has, over time, been
more skeptical than the U.S. approach. This is also reflected in EU case law, which refers to the “special responsibility” of a dominant undertaking
not to allow its behavior to impair genuine, undistorted competition.3

3. Does the consumer welfare standard need to be changed to incorporate public policy considerations in light of the digital econ-
   omy?

I would say that a consumer welfare standard interpreted in a way that takes into account not only price effects but also effects on competition
in relation to other parameters, such as product quality and innovation, can fully cater for the challenges of the digital economy in terms of
competition law enforcement.

       Not to focus solely on price effects will be particularly important in two-sided digital platform markets where sometimes only one side pays
and the other side is provided with a “free” service.

        In such markets one should not readily assume that the side which receives the “free” service cannot be harmed through anticompetitive
behavior in the absence of higher prices. Harm can also result from fewer available choices, a deterioration of product quality or an impact on
distribution or product innovation.

       In this respect, the EU courts have held that in two-sided markets, where the customers in those markets are not substantially the same,
the restrictive effects of a measure in one market cannot be compensated by advantages for the other side, if the measure does not have any
appreciable objective advantages for the first side.4 In other words both sides need to benefit from the efficiencies of the measure in order to
make it compatible with EU competition law.

        A consumer welfare standard which takes into account factors such as impact on choice and innovation appears to be well suited for
digital markets where many services are offered free of charge to consumers. On the other hand, EU competition rules appear less well suited
to pursue policy objectives that go beyond ensuring undistorted competition to the benefit of consumers.

4. The OECD is studying whether the competition toolkit should be adapted to the digital economy. What are your views on this
   topic?

The work being done by the OECD on the toolkit is very much in line with the reflections and studies carried out on digital markets by many com-
petition authorities around the world. Recently Commissioner Vestager decided to set up a panel of advisers from outside the Commission. Their
objective is to seek input on what the key upcoming digital developments are that will affect markets and consumers, and on their implications
for EU competition policy.

       The panel is made up of Professors Heike Schweitzer, Jacques Crémer, and Assistant Professor Yves-Alexandre de Montjoye. They are
working on a report on the future challenges of digitization for EU competition policy, to be delivered by March 31, 2019.

         The Commission will also organize a conference on January 17, 2019 to discuss the topic with a broad variety of contributors. Addi-
tionally, the Commission is seeking contributions in particular from those stakeholders that are involved in or affected by the digitization of the

3 See C‑413/14 P, Intel, para. 135.
4 See C‑382/12 P, MasterCard, para. 242.

                                                                         8                                        CPI Antitrust Chronicle September 2018
Antitrust Chronicle SEPTEMBER VOLUME 3 SUMMER 2018 - COMPETITION POLICY INTERNATIONAL
economy.5 The conference and the report from the Special Advisers are designed to provide input to the Commission’s ongoing reflection process
and to identify the key upcoming digital challenges and their implications for EU competition policy.

5. How do you view the pressure on competition rules to increasingly absorb public policy objectives like privacy or data protection?

Competition rules and data or consumer protection rules are complements and not substitutes. The objective of increasing consumer welfare
overlaps, but different tools are deployed. It would be a disservice to competition law enforcement if one were to expect that its tools can solve
privacy or consumer protection issues meant to be tackled through the enforcement of laws specifically designed for that purpose.

        Instead of overburdening competition law enforcement with pursuing policy objectives it was not meant to cope with in the first place,
it would be wiser to rely on the cooperation of competition law enforcers with data and consumer protection authorities in order to ensure the
complementarity of enforcement activities in digital markets.

6. Which of the many debates around relatively novel issues (like algorithmic collusion, privacy as a parameter of competition, etc.)
   would you consider as most relevant?

One should be careful with predictions, especially about the future. However, one trend that is likely to impact competition law enforcement is
the use of Artificial Intelligence (“AI”) by companies in order to adapt pricing or products. As AI needs (big) data to properly work, it is possible
that disputes will arise between companies about access to valuable data, or companies may merge in order to get access to such data. In these
cases, the assessment under competition rules would likely need to examine the competitive value of data, which is not without challenges for
competition law enforcers.

5 More information is available at http://ec.europa.eu/competition/scp19/.

                                                                             9                                    CPI Antitrust Chronicle September 2018
Antitrust Chronicle SEPTEMBER VOLUME 3 SUMMER 2018 - COMPETITION POLICY INTERNATIONAL
STRENGTHENING BUYER POWER AS A SOLUTION
TO PLATFORM MARKET POWER

BY HUGH MULLAN & NATALIE TIMAN1

1 Natalie Timan is Director of Economics at the UK Competition and Markets Authority (“CMA”). Hugh Mullan is Assistant Director of Economics at the CMA.
Views expressed in this paper are personal and do not represent the CMA.

                                                                         10                                          CPI Antitrust Chronicle September 2018
I. INTRODUCTION
Online platforms play an important role in the development of the online world and open new market opportunities, notably for SMEs. Online
platforms also bring benefits for consumers, for example by reducing search costs and making greater choice available. Almost half (42 percent)
of SME respondents to a recent Eurobarometer survey on online platforms use online marketplaces to sell their products and services. Eighty-two
percent of firms that sell online rely on search engines to promote their products and/or services.2 As the European Commission notes, online
platforms are now the main gateway to market for the majority of smaller businesses in the digital economy – be it online market places for small
sellers, app stores for game developers, or online travel agents for hotels.3

       This led the Commission to identify that many small, but also some larger businesses, have come to depend on platforms that provide
such easy access to customers and markets. The Commission found that this dependency entails a certain imbalance of bargaining power
between online platforms and their business users, causing friction in platform-to-business relationships, and giving scope to unfair behavior
on the part of platforms.4 This may indirectly lead to harm to consumers.5 This reflects the growing concerns about the strong position of digital
platforms.

       A digital platform may acquire a position of market power for a number of reasons. As noted, they tend to act as a key gateway for
businesses engaged in online sales or marketing. In addition, there may be powerful network effects at work. These mean that the value of the
platform to one group of users, say online businesses, increases as the platform gains a greater number of other types of users, here online
consumers, and typically vice versa. These effects will tend to create barriers to entry and favor incumbents, increasing the risk of increased
concentration of markets in which platforms operate.

       The increasing market power of some digital platforms has led to calls for greater regulation.6 Here we discuss whether there may be
more market-orientated solutions to platforms holding very strong market positions, or market power. In particular, we explore whether, and why,
greater collective bargaining by user groups on platforms may temper the power of some digital platforms and lead to better outcomes.

II. EXAMPLES OF COLLECTIVE BUYER POWER FROM THE PROPERTY PORTAL, AUCTION
HOUSE, AND PAYMENT CARD SECTORS
In order to motivate the discussion, we identify some examples where businesses selling through platforms have collectivized to exercise outside
options when dealing with major platforms in what appear to be fairly concentrated markets. In particular, the launch of the “OntheMarket” plat-
form by a collective of estate agents provides an example of how business users might respond to their reliance on powerful incumbent platforms
to reach consumers using platforms to search for properties to buy.7

          OnTheMarket claimed that:8

          a) There were two major incumbents in the UK and estate agents felt obliged to list on both of them to reach consumers. This increased
2 Inception Impact Assessment, http://ec.europa.eu/info/law/better-regulation/initiatives/ares-2017-5222469_en.
3 Ibid.
4 Ibid.
5 The Commission has listed a variety of issues which may arise due to this imbalance in power. The Commission concludes that, “In light of the increasing
dependency of businesses users on online platforms to reach markets, these unresolved issues can have significant direct negative effects on a large number
of EU business users, some of which may choose to disengage from online platforms. Where they hamper the business users’ ability to reach markets, these
issues can also indirectly harm consumers by leading to a more limited choice of products and services. In addition, the prevalence of these issues could have
significant negative effects on the viability of the wider platform business model and the innovation that their ecosystems drive, including on potential new
entrant platforms.” Ibid.
6 Aside from the initiatives of the European Commission, there have been two recent reports from the UK House of Lords suggesting that further regulation of
platforms may be required.
7 OnTheMarket launched in January 2015. In February 2018 it had an IPO and was admitted to the AIM market of the London Stock Exchange. It is the third
biggest property portal in the UK in terms of traffic and competes with Rightmove and Zoopla.
8 See para 69 of the CAT judgment of July 5, 2017, http://www.bailii.org/uk/cases/CAT/2017/15.html#para64.

                                                                            11                                           CPI Antitrust Chronicle September 2018
overall costs to estate agents. OnTheMarket described the incumbents as a duopoly who did not provide a rivalry which would benefit
             consumers and agents. It was claimed that the incumbents had increased listing fees, reduced the quality of service, and imposed
             adverse conditions on listing.

        b) Following an initiative by one agent, a number of agents came together to create “a true competitor” to the incumbents. “Whilst
           barriers to entry are substantial and potentially insurmountable to non-agent ventures, a new portal owned by agents and run with
           the aim of improving service and reducing costs to them, their customers and the property seeking public can win through in the
           medium term.”9

         Another example of buyers coming together to launch a platform is TheAuctionRoom. This is an online bidding platform created by three
auctioneers who had previously used the large platform ATG Media.10 The launch of TheAuctionRoom provided auctioneers with an additional
outside option. However, given that the CMA recently found that ATG Media is likely to have a dominant position in the market for the supply
of live online bidding auction platform services to auction houses in the UK,11 the extent to which the launch of the new platform improved the
bargaining position of auctioneers is less clear cut. This illustrates that launching an alternative platform successfully is challenging and may not,
therefore, provide a strong outside option when negotiating with an incumbent platform.

        A final example of buyers coming together to exercise buyer power in relation to powerful platforms was seen in the context of debit cards.
There are two major (four party) schemes for debit and credit cards, Mastercard and Visa. The CAT recently assessed the ability of merchants
to influence the interchange fees charged by issuing banks under the schemes. In doing so, it noted an example of collective bargaining by
retailers. The CAT pointed to the British Retail Consortium, a trade association representing retailers, successfully rejecting an ad-valorem pricing
structure proposed by Mastercard.12 Retailers were able to reject Mastercard’s offer and, instead, go with the other major debit card provider,
Visa. This illustrates how collective bargaining may be used to choose between existing platforms, which may generally be a more attractive way
of exercising bargaining power than launching a new platform.

III. AN APPLICATION TO HOTELS AND ONLINE TRAVEL AGENTS
To further motivate the discussion which follows, we identify an example of an industry where there appear to be powerful platforms and no
collective bargaining. In particular, we consider the service supplied to hotels by online travel agents, such as Booking.com and Expedia.

        This is an interesting sector because: (i) it is highly concentrated, with Booking.com and Expedia (and the sites owned by the same cor-
porate groups) tending to account for a large proportion of bookings on OTAs; 13 (ii) there has been a lot of recent antitrust scrutiny of this sector
as well as consumer law concerns; and (iii) the hotel industry is a fragmented sector and, even though there are some large hotel groups, they
account for a very small proportion of all hotels. Therefore, this sector might provide a good example of where the bargaining power of individual
hotels is rather weak relative to the bargaining power of the OTAs.14

9 By coming together to create a new competitor, sellers were able to create an additional outside option, which in turn, should have helped their bargaining
power.
10 The Auction Room’s website claims it is a “new search engine hub, portal and auction platform uniquely created by three of the top UK auctioneers for their
buyers, with aspirations to be the best for live auction bidding in the UK. It is designed to deliver the experience of the true auction room atmosphere.” https://
www.theauctionroom.com/aboutUs.
11 See paragraph 3.2 of the CMA’s decision to accept binding commitments offered by ATG Media in relation to live online bidding auction platform services,
June 2017 https://assets.publishing.service.gov.uk/media/5954be5c40f0b60a44000092/auction-services-commitments-decision.pdf.
12 See paragraph 261 of the CAT’s Judgment of July 14, 2016, in Sainsburys v. Mastercard.
13 It has been estimated that Booking.com and Expedia had respective shares of 44 percent and 18 percent of bookings through online travel agents in Europe
in 2016 (European Online Travel Overview, 12the Edition). OTAs account for about 60 percent of online bookings for hotels and 40 percent of all hotel bookings
in the EU. Table 1. Report on the monitoring exercise carried out in the online hotel booking sector by EU competition authorities in 2016. http://ec.europa.eu/
competition/ecn/hotel_monitoring_report_en.pdf. We note that these figures only provide a snapshot of concentration amongst OTAs and do not seek here to
make any assessment of the dynamics of this industry.
14 The British Hospitality Association summarized the situation from its perspective as follows: “it is no longer feasible for the majority of hospitality and tourism
venues to operate without online sales, as most searches for hotel, restaurant and other bookings are made online. Venues must therefore make agreements
with online platforms and are obliged to choose the largest platforms, since these are the ones visited by customers. They are visible on the first pages of a
customer’s search, venues own websites appearing several pages on. As a result, venues, in particular SMEs have little bargaining power and agreement terms

                                                                                12                                             CPI Antitrust Chronicle September 2018
Hotels face a coordination issue. An individual hotel could threaten to leave one OTA (e.g. Booking.com) and offer their rooms only through
the other OTA (e.g. Expedia) unless the hotel was offered a better rate by Booking.com. However, if the hotel does offer its rooms only through
Expedia, then it will lose those customers who search for hotels through Booking and do not use Expedia, or the hotel’s own website.15 Therefore,
the threat to switch may not be credible. In addition, hotels also face a size problem – any hotel or hotel group (even the largest groups) are
small relative to total hotel bookings on an OTA. This means that an OTA may have little incentive to adjust its commission rate significantly to
keep that hotel group.

          If hotels could coordinate their purchases from OTAs, they would be better placed to credibly threaten to switch. If all, or a large proportion
of, popular hotels coordinated to leave a particular OTA, then consumers who previously used only that platform would face a strong incentive
to also switch OTA (or go directly to the hotels’ own websites or possibly search across multiple OTAs). A loss of hotels on the platform, which
in turn leads to a loss of consumers, would make it easier for more hotels to switch away from the platform (and less painful for the group that
initially left), which would lead to yet more consumers leaving. This could be very damaging to the OTA. Therefore, the threat may be credible
and powerful.

        A credible threat to switch would strengthen hotels’ bargaining power, and, in turn would allow hotels to push down commission rates
(and/or secure better standards of service). To the extent that the hotel sector is competitive,16 a reduction in commission would likely be passed
through to some degree to consumers (either through lower prices as one would expect that commission rates are included in the price of hotel
rooms, or in other improvements to the offering). The commission rates of the major OTAs range from 10 percent to 20 percent of the price of
a hotel room.17 Therefore, where the downstream market in which the sellers operate is competitive, the potential benefits to consumers could
be great.

are controlled by the online platforms, resulting in contracts which include ‘parity clauses’ and ‘most favoured nations’ clauses. A myriad of other concerns
arise as a result of this platform/industry/customer imbalance. These include excessive commissions and lack of transparency on the part of online platforms,
whose websites are often confusing and misleading for the customer, for example, in not providing the customer with key information about pricing, how ratings
are awarded to venues, terms of booking and so on. The need for answers to these problems is pressing as the most powerful online platforms have grown in
strength very rapidly, in control of the market and in wealth, giving them enormous resources to consolidate their position.” BHA written evidence to the House
of Lords EU Internal Market Sub-committee.
15 Customers that only use one OTA are said to “single-home,” whereas customers that use more than one OTA are said to “multi-home.” Hotels may need to
multi-home in order to reach those consumers than single-home.
16 We have made no assessment in this article as to how competitive the hotel sector is. We simply note that it is fragmented and therefore provides a useful
illustrative example.
17 In addition, the commission rates charged by OTAs have not fallen in recent years, despite the removal of wide price parity terms from OTA contracts with
hotels which were believed, by competition authorities, to have sustained supra-competitive commission rates. Page 18. Report on the monitoring exercise
carried out in the online hotel booking sector by EU competition authorities in 2016. http://ec.europa.eu/competition/ecn/hotel_monitoring_report_en.pdf.

                                                                             13                                           CPI Antitrust Chronicle September 2018
IV. THE DETERMINANTS OF BARGAINING POWER18
The availability of credible alternatives (or outside options) is a key factor in determining which parties will hold a strong bargaining position.19
In its market study on Digital Comparison Tools (which are digital platforms), the CMA identified three factors which affect the strength of a
platform’s bargaining position vis-à-vis the sellers on this platform: (i) the significance of platforms overall as a sales channel for these sellers;
(ii) the significance of an individual platform to one or more sellers; and (iii) sellers’ ability to replicate the sales they make on the platform on
other sales channels, such as their own websites.20 A seller’s threat to delist from a platform will be less credible if it is more dependent on that
platform and, therefore, the platform will hold a stronger bargaining position.21

          A seller’s position will be stronger if they are able to replicate sales on other channels (including other platforms, or the seller’s own web-
site), if they were to delist from the platform, or steer consumers to these alternatives. This, in turn, will depend on the extent to which consumers
are willing to use these other channels and how cost-effective these other channels are.

        For a given structure in the supply of platform services, a seller will tend to be in a stronger negotiating position the stronger its brand is in
the eyes of consumers and the larger its share of supply.22 Similarly, the threat of switching away to another platform may also be more credible
(and, clearly, have a far greater impact on the current upstream supplier) when buyers can coordinate on whom they purchase from. In this way,
discounts may also be achieved through more active negotiation by the buyer group. The accretion of buyer power, through greater concentration,
is often recognized as a potential benefit.23 Therefore, if the platform holds a strong bargaining position and a position of market power, collective
bargaining by businesses using that platform may lead to benefits to both those businesses and the customers to whom those businesses sell.

V. PLATFORM MARKET POWER AND THE POTENTIAL BENEFITS OF COLLECTIVE
BARGAINING
As explained above, the bargaining strength of any customer will depend on the number and strength of the credible outside options available to
that customer. A specific challenge for sellers on platforms in exercising outside options and achieving greater bargaining power vis-à-vis plat-
forms arises from the multi-sided nature of platforms, which display indirect network effects. These effects mean that for a platform to succeed
it must attract both sides of the market. These effects will also tend to strengthen the position of established platforms with loyal consumers.

18 This article focuses on a type of countervailing buyer power that buyers might achieve through collective bargaining. It does not examine buyer power
stemming from the aggregation of purchases from a number of buyers which may lead to reductions in the costs of supply of the upstream supplier (or platform)
due to the realization of economies of scale, leading to discounts on larger volumes being purchased.
19 The CMA noted that “a significant factor affecting prices paid by consumers on DCTs is the outcome of the negotiation between DCTs and suppliers. This
depends primarily on whether the DCT and the supplier have credible alternatives they can exercise if an agreement is not favorable to that party. This is closely
linked to how strongly DCTs compete with each other (and suppliers’ direct channels) and to the strength of competition between suppliers.” See Paragraph
2.37, Digital Comparison tools market study, Paper E: competitive landscape and effectiveness of competition. https://assets.publishing.service.gov.uk/me-
dia/59e093f5e5274a11ac1c4970/paper-e-competitive-landscape.pdf.
20 Paragraph 2.42, Ibid.
21 For example, the CMA found that platforms were a more important sales channel in motor and home insurance than in broadband and credit cards and that
this was consistent with commissions following an upward trend in motor and home insurance. The CMA also found evidence of high levels of concentration of
platforms within some sectors and high levels of consumer brand loyalty, both of which would strengthen the negotiating position of the platform.
22 As the CMA noted, “In markets with a low concentration of suppliers, we would generally expect DCTs to have a stronger negotiating position because there
are potentially many other suppliers a DCT can contract with. In addition, where suppliers are relatively undifferentiated, this is likely to strengthen a DCT’s
negotiating position.” See Paragraph 2.39, CMA market study on DCTs, Paper E.
23 For example, the CMA’s Merger Guidelines note how greater buyer power due to the merger may lead to benefits to the customers of the merged firm.
“Where the merger firms purchase the same products, the merged firm may enjoy greater buyer power (or monopsony power) than the merger firms could
previously exert individually. In many cases, an increase in buyer power is not likely to give rise to unilateral effects; and some of the benefits to the firm from
its greater buyer power may be passed on to the merged firm’s customers.” Paragraph 5.4.19, Merger Assessment Guidelines. https://www.gov.uk/govern-
ment/publications/merger-assessment-guidelines. The European Commission’s guidelines on horizontal agreements also recognize the benefits of collective
purchasing to strengthen buyer power. See paragraph 194 https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52011XC0114(04)&from=EN.

                                                                               14                                             CPI Antitrust Chronicle September 2018
If consumers are content to search across different platforms (i.e. multi-home), then the seller would only need to list on one platform –
safe in the knowledge that these consumers will search sufficiently to find their offer.24 In this context, an individual seller’s threat to switch from
a platform may be credible, and collective bargaining may not be very important in strengthening their credibility.

         Sellers can only credibly threaten to switch or delist from one platform if they can also steer consumers to the alternative platform. This
is more difficult if consumers single home (i.e. use only one platform, rather than searching across multiple/all platforms). If these single-homing
consumers represent an important source of demand for sellers, to whom they would not be able to appeal to if not on the platform, the sellers
may consider it necessary to sell through a platform to reach these consumers. This may give the platform a degree of market/bargaining power
in relation to the seller, as the latter cannot switch away from the former if it wants to attract these consumers. Therefore, the sellers are likely to
multi-home (i.e. choose to list a given product/service across multiple platforms) due to the single-homing behavior of some significant groups
of consumers. Although there may be multiple platforms offering a similar service, each platform may have market/bargaining power in relation
to the sellers on that platform due to single-homing by an important group of consumers.25

        This provides sellers with a coordination problem. A single-homing consumer who has built up a preference for using a particular platform
is unlikely to switch to another platform in response to only a few sellers switching to that platform. Therefore, in order for sellers to exercise bar-
gaining power through a credible threat to switch, they may need to coordinate amongst themselves such that a sufficient number of important
sellers26 switch, providing sufficient incentives for the consumers to also switch.27

        Sellers could coordinate to threaten to switch to an existing platform, or to effectively sponsor entry by creating a new platform. Once the
threat to switch to an alternative becomes credible, the balance of power should shift away from the incumbent platform, which in turn could lead
to lower commissions for the seller.28 Provided that the sellers pass these cost savings on to consumers and do not coordinate on any aspect of
competition between them on the selling market, this collective bargaining could benefit both sellers and consumers.

        The nature of indirect network externalities means that it is more challenging for sellers to credibly threaten to switch. On the other hand,
if a credible threat exists, the exercise of this threat could be particularly damaging to the platform. A sufficiently large switch by sellers may
cause the platform to become unsustainable and go into a “death-spiral”: as switching by sellers induces switching by consumers, this induces
further switching by sellers which induces further switching by consumers, and so on until the platform no longer has sufficient business to be
sustainable.29 In addition, this dynamic could also mean that those sellers who switched are left with no other option to whom to switch in future,
having effectively shifted market power from one platform to another, at least over the long run. Such concerns may not arise in reality: not all
sellers may wish to switch, so a death-spiral might not occur.

        Furthermore, it is possible that switching consumers from one platform to another would transform the behavior of these consumers from
being single-homing consumers to being multi-homing consumers. These consumers may then, as part of their normal search activity, consider
offers across multiple-platforms. So long as search costs are not significantly increased, and benefits are passed onto consumers, this would
be a positive outcome. It would mean the coexistence of two or more platforms is more likely, as it weakens the indirect network externalities. In
doing so, it would reduce both the coordination problem faced by sellers (in response to single-homing consumers) and also remove the threat
of a death spiral for platforms. Whether consumer behavior would actually be transformed in this way, over the longer term, is unknown.
24 This is a somewhat strict definition of multi-homing – consumers searching all platforms. In reality, some consumers may search more than one platform,
but may not search all available platforms. Nevertheless, the arguments made here, regarding multi-homing consumers, are likely still to hold true (to a greater
or lesser extent) if consumers visit multiple platforms.
25 As the CMA noted, “given that, in practice, a significant proportion of consumers single home, most suppliers choose to use multiple DCTs in order to get
access to as many consumers as possible.” See Paragraph 2.5, CMA market study on DCTs, Paper E.
26 The actual number of sellers switching need not be large if they are sufficiently differentiated and important in the eyes of consumers that they can affect
the attractiveness of a platform.
27 Sellers multi-homing on platforms (in response to single-homing by groups of consumers) means that each seller is already on each platform. Switching
does not necessarily mean moving to a new platform on which the seller was not previously listed. Rather, switching may mean no longer selling through a
platform on which a seller was previously listed; instead, concentrating sales through an alternative platform.
28 Single-homing consumers suggests a lot of consumer inertia or resistance to switching from their current platform. A strategy to shift away from the platform
only works if sellers can minimize the impact/overcome consumer inertia, which may mean some loss of profit or extra cost in the short term. The higher the
proportion of sellers switching, the more these costs can be spread. In addition, the greater is the likelihood of switching consumers (and reducing these costs).
29 Indeed, the issue of such death-spirals has pre-occupied, and continues to pre-occupy, competition assessments of credit card interchange fees due to the
network externalities arising there. For example, see the Court of Appeal Judgment, July 4, 2018.

                                                                              15                                            CPI Antitrust Chronicle September 2018
The key point of this section is that collective bargaining may play an even more important role in relation to platforms than in other types
of markets, due to network effects and the single-homing nature of many consumers. This is because the sellers on the platform need to ensure
that they can bring the consumer side of the market along with them in order to have a credible threat of switching and to exercise some degree
of bargaining power.

VI. POTENTIAL CONCERNS ABOUT GREATER COLLECTIVE BARGAINING BY USERS OF
PLATFORMS
As explained above, collective bargaining can in certain circumstances benefit consumers and businesses that use digital platforms. However, it
is important that when businesses come together to solve one type of coordination problem (such as not being able to switch to an alternative
platform), this does not lead to coordination over any feature of competition between the businesses. Here, we discuss instances where greater
collective bargaining by platform users may lead to consumer harm.

        These theories of harm may include the following:30

        a) Cooperation in the purchasing market may facilitate collusion in the selling market, which would be directly harmful to customers.
           There could be collusion on selling without cooperation in buying, but the latter could lead to the former. This might arise because,
           say, establishing the buyer group increases communication or transparency between the sellers on a platform. In particular, cooper-
           ation that leads to the sharing of private information could facilitate collusion in the selling market.

        b) Cooperation by a subset of buyers may allow strategic behavior that harms the terms of supply for other buyers, which are not part of
           the group of cooperating buyers. This would, therefore, raise the costs of rivals and may weaken competition in the market in which
           the collective-buyers sell.

        c)   Platforms increase price competition by allowing consumers to search and compare across many providers selling through a plat-
             form.31 If sellers were acting collectively when purchasing services from the platform, there may also be a collective realization that
             they would be better-off if all sellers chose no longer to list on the platform and thereby avoid the intensity of competition which listing
             on a platform may entail. Consumers would be made worse-off in the absence of platforms because they may face higher prices and
             would lose the ability to search and compare easily.

        d) Cooperation in the purchasing market which lowers the price paid to the input suppliers may lead to the supplier being able to invest
           less in developing its products and, more generally, innovating.

        The likelihood of harm in any case will depend on the specific facts of the case. However, these theories of harm are less likely to be
significant concerns when the seller side is fragmented and more so when the platform side is concentrated and characterized by strong sin-
gle-homing tendencies by many consumers. Indeed, we do not think that these theories of harm should lead one to ignore the potential benefits
of collectivization of buyers in order to exercise buyer power. Buyer groups have always existed and have traditionally led to few concerns from
competition authorities.32 Moreover, we consider that there may be material benefits to customers if there is greater collectivization of buyers in
response to significant market power in the input markets from which they purchase.

      It is worth noting that it is not always clear who the (downstream) buyer is when assessing the commercial dynamics of platforms. For
example, a hotel may be a purchaser of platform services from a powerful online travel agent, where collective bargaining might lead to an
outcome that is more beneficial for consumers. Alternatively, one might characterize a hotel as a supplier of hotel rooms for sale through a
30 There are a number of other potential harms from buyer power, but they do not appear to be relevant when applied to users of platforms, particularly when
the platform has a strong degree of bargaining power or market power.
31 The CMA found evidence in both its Digital Comparison Tools market study and its Private Motor Insurance market investigation that the price sensitivity
of consumers on comparison sites is considerably higher than on other sales channels, which puts competitive pressure on the sellers listing on those sites.
32 It is worth noting that the ACCC’s recent discussion paper considers a class exemption to allow small businesses and franchisees to collectively bargain
without prior consent from the competition authority. The ACCC notes that it has “generally not had concerns about collective bargaining applications by groups
of small businesses, but collective bargaining by larger businesses has greater potential to raise competition concerns (for example, by reducing competition
and leading to consumers paying prices above competitive levels). For this reason, our starting point is that a class exemption for collective bargaining should
only be available to businesses below a certain size.” https://res.cloudinary.com/gcr-usa/image/upload/v1535039648/ACCC_class_exemption_dclioa.pdf.

                                                                             16                                            CPI Antitrust Chronicle September 2018
(downstream) platform. In this scenario, collective bargaining by hotels might appear more like a seller cartel, agreeing on the terms of supply.
This illustrates how upstream and downstream, buyer and seller, may not be obvious in relation to multi-sided platform markets. However, from
an economics perspective this should not really matter. What matters is that the collectivization, which leads to increased negotiating power for
the parties selling through the platform, is likely to bring net benefits to consumers. Tolerance of greater collectivization of groups using powerful
platforms may well achieve these benefits.

VII. THE ISSUE AND OUR RECOMMENDATION
Online platforms bring clear benefits to their customers, for example, by opening routes to consumers that SMEs might not have otherwise had
access to and by providing more choice to consumers with reduced search costs. Nevertheless, there is a sense that, in some sectors, digital
platforms are gaining increasing power and that this places the third parties which use them at a significant disadvantage. This may lead to
harm to consumers, either indirectly through the treatment of the businesses using the platforms or more directly through the exercise of market
power in terms of the commissions and prices platforms charge to the businesses selling/advertising through them, which will be passed on to
consumers.

       One way of tempering stronger market power in platform markets is through increased buyer power on behalf of the businesses selling
through a platform with market power. We have offered some examples of where this has been used before, whether through the use of trade
associations as a collective negotiator or through sponsored entry on an alternative platform by businesses selling through platforms.

        However, it also strikes us as surprising how infrequently this collective action seems to take place given the apparent strength of some
platforms. This may well be down to the nature of the agreements with platforms, which are individually negotiated and often long term. However,
it may also be down to the differing objectives of buyers, such that it is too hard to coordinate to establish a stable buyer group and negotiate
an outcome beneficial to all buyers. This is likely to be the case particularly when there is diversity in the size of the businesses and when ne-
gotiations relate to commercially sensitive input prices, such as commission rates. When these businesses then compete with each other quite
directly, the incentive to negotiate collectively may also be reduced.

       Another reason why we may not see a lot of collective bargaining with platforms may be because of competition law concerns about busi-
nesses who are competitors cooperating as buyers. We have outlined a number of theories of harm which may arise in this respect, in particular
where such cooperation would facilitate or support anti-competitive collusion. However, there is a distinction between these two practices and
businesses will be aware of the differences (as illustrated by the long existence of legal buyer groups). Another possible concern is that cooper-
ation by businesses selling through platforms could lead to reduced use of platforms or long-term damage to the platform model.

        We think that collective bargaining by sellers using platforms may provide a beneficial counter-balance in cases where platforms hold a
particularly strong position. Moreover, while greater regulation of platforms appears to be gaining greater support, it seems to us that in certain
circumstances, increased collective bargaining by sellers in the presence of platform market power may provide an alternative, more market-ori-
entated, response to such concerns.

                                                                         17                                        CPI Antitrust Chronicle September 2018
You can also read